Context awareness for content delivery over mobile networks

The last few years has seen an enormous growth in the usage of smart phones. This trend is expected to continue in the near future as well with the availability of more processing power and larger form factors. This explosive growth has contributed to an unprecedented adoption of mobile data services which results in severe network congestion during peak hours due to limited availability of cellular capacity and the prohibitive cost of installing new infrastructure. Even the recent adoption of Long Term Evolution (LTE) by carriers has not kept pace with increasing data needs and the networks continue to be under stress. However, it is of utmost importance for the provider to maintain superior Quality of Experience (QoE) and Quality of Service (QoS) even under constrained network capacity. One possible approach for achieving high QoE is through the use of context information from smart phones that allow the network to take pro-active decisions for content delivery.;Key players within the telecommunication industry seem to be unanimous in foreseeing, for the next few years that the Machine-to-Machine (M2M) traffic trends will grow rapidly. The Applications on mobiles, serving content, can be considered the present manifestation of the growing trend. Understanding the potential impact of M2M traffic on current network dimensioning and architectures is of paramount importance for the success and sustainability of these novel services in future wireless mobile networks.;This thesis presents a tool "Melange" which can be used to measure and demonstrate the benefits of using context awareness for content delivery. Using the tool, the user can remotely control a number of mobile handsets by deploying different types of synthetic and natural workloads, emulating the traffic, to evaluate the impact on battery consumption and user perception of context-aware content pre-fetching.